1. PM with AI Scheduling
  2. Features
  3. Operations oversight
Feature · Operations oversight

Operations oversight

Use receipts and the overview to trace exceptions before closing work.

OverviewOperations needing review
PM with AI Scheduling · Operations oversight

Operations needing review

Operation receipts

New
ParentOperation keyActionExpected revisionResult stateOccurred at
WP-27976OK-788ACT-12650Failed03 Oct
WP-31740OK-521ACT-643550Needs review19 Oct
WP-59709OK-831ACT-52450Failed22 Oct
WP-81180OK-349ACT-878590Needs review11 Oct
WP-49584OK-942ACT-114280Failed27 Oct
6 of 6 records
OK-788Operation receipts · OR-61935
Failed
ParentWP-27976
ActionACT-126
Expected revision50
Occurred at03 Oct
Failure reasonCut to length
Example screen with sample data

What you can do

Operations oversight

Operation receipts record a guarded action, key, expected revision, result state, occurrence time and any failure reason or linked effect references. They let the delivery lead separate a completed action from one that failed or needs review. Operations Overview brings together active and accepted work, pending decisions, failed operations, planned and actual task hours, and overdue open tasks by owner. Source links lead back to the affected package, decision, task or receipt.

How the work is organized

  • Exception queue. Open the package and receipt behind each failed or uncertain result.
  • Hour comparison. Review planned and actual task time in project context.
  • Completion check. Verify linked effects before closing an outstanding receipt.

What to agree during setup

Agree who investigates failed operations and who can resolve a needs-review result. Confirm what counts as an observed effect for each guarded change. The lead should keep the original operation key when checking a retry, so the team does not confuse a repeated request with a new action.

Check the handoff on an actual case

A coordinator attempts to apply a reviewed schedule, but the operation receipt shows needs review. The delivery lead opens the linked package, checks the expected revision and task lines, and compares the recorded effect references with the actual work. The lead records the outcome only after confirming the effect. Meanwhile the overview still shows the package and exception in context, allowing the team to continue investigating without counting the action as completed.

Review the exception queue as part of the normal delivery meeting. For each receipt, start with the action and parent package, then compare the expected revision with the package now on screen. The effect references point to records that should have changed. Open those records before marking a result complete. If they do not match, keep the exception visible with its failure reason while the coordinator investigates. The same discipline helps with schedule approvals and final acceptance: the lead sees a pending decision as pending, rather than treating a proposed outcome as finished work. Planned and actual task hours in the overview help identify the package that needs attention next.

What your team sees

  • Operations needing review — Investigate failed and uncertain receipts with their package context.
  • Active work packages — Check delivery status alongside exceptions.

Follow the work through

Follow operations oversight in the operating process. Review the data model when mapping project, package and evidence records.

What it keeps track of

Data model
Operation receipts
Durable receipts for every guarded operation, keyed for idempotent retries.
Work packages
Contracted units of delivery with lifecycle, budget hours and accepted baseline.
Tasks
Scheduled task lines inside a work package with planned and actual hours.
Scheduling proposals
AI-generated schedule proposals awaiting review and approval for a work package.
Review decisions
Independent reviewer decisions recorded against a work package revision.

Who uses it

Permissions
Delivery Lead
Owns the exception queue; Project Coordinator supplies operation context; Admin supports access and identity checks.

Set up PM with AI Scheduling your way.

Start free, then use Proto to change fields, steps and approvals before your team goes live.

How to get started